Series 3 caller ID: NCID

[Soapm]

I have this entry in my cron file set to run daily /var/hack/bin/startncid rmpid

However it seems my modem stays locked up and I get these errors.

127.0.0.1:3333 - error reading "sock15": connection refused

A reboot always solves the problem.

I notice I have this file which I assume should reinitialize my modem /enhancements/varhacks/hack/bin/initmodem but I don't know how to configure it or if it will work. Will this file work for me or is there another way to reinitialize my modem without rebooting the Tivo?

I would think if I can't reinitialize my modem then I should remove the cron entry and just let ncidd run until it stops.

TivoHD with latest OS...

PS... is there a command I can put in my cron file that would manually reinitialize or at least unlock my modem? Perhaps I can set that to run just before the NCID reset.

[jlc]

I have this entry in my cron file set to run daily /var/hack/bin/startncid rmpid

However it seems my modem stays locked up and I get these errors.

127.0.0.1:3333 - error reading "sock15": connection refused

A reboot always solves the problem.

The error indicates ncidd is not running anymore. The /var/hack/bin/startncid script is not safe when trying to start programs already running. You would need to kill ncidd and ncid (either tivocid or tivoncid) before running the script.

I notice I have this file which I assume should reinitialize my modem /enhancements/varhacks/hack/bin/initmodem but I don't know how to configure it or if it will work. Will this file work for me or is there another way to reinitialize my modem without rebooting the Tivo?

I would think if I can't reinitialize my modem then I should remove the cron entry and just let ncidd run until it stops.

TivoHD with latest OS...

PS... is there a command I can put in my cron file that would manually reinitialize or at least unlock my modem? Perhaps I can set that to run just before the NCID reset.

You can put /var/hack/bin/initmodem in a cron file and it should reinitialize your modem.

You and also use the initmodem client module, edit the /var/hack/bin/startncid script, and remove the '#' from the line:

#INITMOD=ncid-initmodem

The ncid-initmodule module will notice the modem is not in CID mode during a call and should reinitialize it. NCID will miss that call but it will capture future calls. You can use this module and the initmodem script together. They will not interfere with one another.

[Soapm]

The error indicates ncidd is not running anymore. The /var/hack/bin/startncid script is not safe when trying to start programs already running. You would need to kill ncidd and ncid (either tivocid or tivoncid) before running the script..

Thanks for responding... I thought the script first used pgrep to locate the PID's and kill any running portions of NCID before restarting it. That is why I made it a cron job to run each day. If it doesn't kill the PID's first then I guess there's no point it trying to have it run each day. I thought I had something like this running on my old S2DT but I guess I didn't.

The problem I have is NCID will run a few days then a call will come and nothing shows on the screen. I can kill all the PID's and restart it and it will run a little while longer. I thought I could eliminate having to manually restart the app if I have it "refresh" itself each day.

[jlc]

Thanks for responding... I thought the script first used pgrep to locate the PID's and kill any running portions of NCID before restarting it. That is why I made it a cron job to run each day. If it doesn't kill the PID's first then I guess there's no point it trying to have it run each day. I thought I had something like this running on my old S2DT but I guess I didn't.

You are correct if you are using a recient version of NCID. I assumed you were using an old version of the startncid script and then completely forgot about the newer version using pgrep. Here are the first few lines of the current version:

#!/bin/sh
# script to start NCID
# Requires the "pgrep" command
# Last modified  by jlc: Mon Aug 30, 2010

If you are using a current version of startncid, then make sure it starts ncidd. It could be killing it and not restarting it. Look at the end of the /var/log/ncidd.log to see if it was killed and not restarted or if it got into trouble when restarting.

The problem I have is NCID will run a few days then a call will come and nothing shows on the screen. I can kill all the PID's and restart it and it will run a little while longer. I thought I could eliminate having to manually restart the app if I have it "refresh" itself each day.

The /var/hack/bin/initmodem shell script should be used as a cron job if you think the modem is a problem. It uses ncidd to reinitialize the modem. You can run it manually and then check ncidd.log to make sure it initialized the modem. It was last modified in 2010.

[Soapm]

I'm still having trouble keeping ncid running. I was looking at the script startncid and it looks like it kills the running ncid with this line of code.

[ "$1" = "rmpid" ] && rm -f /var/run/*ncid*.pid

My ncid is running and I looked in /var/run and see only klogd.pid syslogd.pid. Could this be stopping scartncid from killing my running version? What can I do so it can kill the version running thereby restarting it per my cron entry? using ps aux I can see it's running on 490 ncidd however since my current pids are above 20K I can tell that ncid isn't being stopped and restarted each day.

[jlc]

I'm still having trouble keeping ncid running. I was looking at the script startncid and it looks like it kills the running ncid with this line of code.

[ "$1" = "rmpid" ] && rm -f /var/run/*ncid*.pid

My ncid is running and I looked in /var/run and see only klogd.pid syslogd.pid. Could this be stopping scartncid from killing my running version? What can I do so it can kill the version running thereby restarting it per my cron entry? using ps aux I can see it's running on 490 ncidd however since my current pids are above 20K I can tell that ncid isn't being stopped and restarted each day.

For a while the various NCID programs always created PID files (the yac2ncid gateway still does because it uses it). Currently PID files are options. The startncid script was thrown together to make it easy to start the various programs at the time when PID files were always created automatically. The code above makes sure old PID files were removed before anything was started. It does not kill off any running programs. The way startncid currently starts the programs no PID files (except for yac2ncid) are generated so the above line is not needed, but I think startncid needs to be modified to create the PID files in the next release. Then startncid could also be modified to stop them and could be used in crontab.

Last time I looked at the problem where the ncidd server stopped working, it was the modem that quit sending CID information to ncidd. When that happens you either need to restart ncidd which configures the modem for CID at startup or just have ncidd reconfigure the modem without restarting. If you run the initmodem script from a cron entry, it should fix the problem by having ncidd reconfigure the modem for Caller ID.

If the ncidd server does not see a caller ID by ring number 2, it defaults to generating a Caller ID using "RING" for the number and "No Caller ID" for the name. The ncidd.log file should contain this line: Handles modem calls without Caller ID

If you enable the Initmodem Client Module in startncid, it will run ncid-initmodem which will reconfigure the modem each time it sees "RING" for the telephone number. Note that this module only fixes the Caller ID modem output for the following calls and may be useful for troubleshooting since the date and time of the generic Caller ID will be noted in the cidcall.log file. You can run both the initmodem script and the ncid-initmodem client module without problems.

[rcflyer]

I got NCID working today on a Tivo HD but have a strange problem. After call clears @stdout is left on the screen. I'm using an alias file and it displays the caller correctly. I start NCID in rc.sysinit.author with /var/hack/bin/startncid rmpid. Any ideas?

[jlc]

I got NCID working today on a Tivo HD but have a strange problem. After call clears @stdout is left on the screen. I'm using an alias file and it displays the caller correctly. I start NCID in rc.sysinit.author with /var/hack/bin/startncid rmpid. Any ideas?

Is the "@stdout" displayed with the Caller ID? Or is it displayed after it is cleared?

It would be helpful if you told me what version of NCID you are using and what programs. I'll assume you are using tivoncid for the client which uses ncid and the ncid-tivo module.

The module uses /tvbin/text2osd to write and clear the screen. You can run it directly and test its output. The following code uses text2osd the same way the ncid-tivo module does. The first line displays the message and the second line clears it:

echo -e "This is a test" | /tvbin/text2osd --line 1 --xscale 2 --yscale 2
/tvbin/text2osd --clear

If the above works as expected, you can add echo statements in the ncid-tivo shell script.that redirect to a file in /var/tmp. Then you can inspect any and all variables.

[rcflyer]

@stdout shows after the call id clears. I can clear it with any button on Tivo remote. I'm using ncid 0.83-1. I am using tivoncid and I have tested with text2osd and it displays fine. I'm expanding to larger drive right now, so I won't be able to do any further testing until tomorrow. Thanks for the quick reply.

[jlc]

@stdout shows after the call id clears. I can clear it with any button on Tivo remote. I'm using ncid 0.83-1. I am using tivoncid and I have tested with text2osd and it displays fine. I'm expanding to larger drive right now, so I won't be able to do any further testing until tomorrow. Thanks for the quick reply.

If it shows up after it clears the screen, then maybe the way it uses text2osd to clear the screen is causing it.

The last four lines of tivo-ncid are:

sleep $TivoDelay
$TivoOSD $TivoCS

exit 0

Try modifying the $TivoOSD line to something like:

echo | $TivoOSD $TivoCS

[rcflyer]

The file is ncid-tivo and that fixed the problem. I did have a reboot when watching Netflix the other night. The caller id came up, then the Tivo menu, then it rebooted. I'll try again and see if I can re-produce it. Thanks again for the fast response and the fix John.

John

[jlc]

The file is ncid-tivo and that fixed the problem. I did have a reboot when watching Netflix the other night. The caller id came up, then the Tivo menu, then it rebooted. I'll try again and see if I can re-produce it. Thanks again for the fast response and the fix John.

John

It's a good thing my mixup of the module name did not confuse you

The fix is simple enough it that is should not break older TiVo boxes so it will be in release 0.84, which will be very soon.

Hopefully your reboot was not caused by text2osd.

[rcflyer]

It's a good thing my mixup of the module name did not confuse you

The fix is simple enough it that is should not break older TiVo boxes so it will be in release 0.84, which will be very soon.

Hopefully your reboot was not caused by text2osd.

Reboot with Netflix is repeatable. I will look for the next release. Hopefully that will take care of it. I love the app. It is so nice to not have to look at a phone when someone calls.

[jkozee]

I am pretty sure the reboot you are seeing is caused by using text2osd on an HME application. I have seen reboots when using text2osd from the command line. I think it may even happen when you are just browsing the HME list (not running an app), but would have to confirm that. IIRC it wasn't an instant reboot. Soemthing to try is use text2osd from command line "text2osd --line 3 --message test" while using Netflx and see if it reboots and how long it takes. If it does reboot, try writing to the screen with text2osd, wait a second and call text2osd --clear and see if it still reboots.

[jkozee]

I can confirm that using text2osd and osdwriter when other HME apps are running will cause a reboot. Seems to reboot as soon as I use the remote. I don't see an workaround, short of issuing a LiveTV remote command prior to calling text2osd, or trying to determine if the UI is currently in an HME prior to calling it.